JPH083946B2 - Servo gain compensator - Google Patents

Description

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention generally relates to a magnetic head positioning system in a magnetic disk drive, and more particularly, the servo of the disk drive is connected to different magnetic heads, and the influence of changes caused by the medium is described. The present invention relates to a servo gain compensator for automatically compensating for a change in servo gain when receiving.

[Technical background of the invention and its problems]

Many disk drives use magnetic heads,
In the track following mode, the recorded servo code is used to keep the head at the center of the track during the read / write operation. The electromagnetic transducers of these magnetic heads consist of a magnetic circuit with winding coils.

The effective magnetic width of these magnetic circuits changes due to their design and manufacturing process. Since this change is not constant for each magnetic head, the servo gain changes when they are individually connected in the servo loop. In some past examples, manual adjustments of servo gain were made to compensate for these magnetic head changes.

Recently, an automatic gain control system for disk drives has been invented to automatically compensate for servo gain changes. At the present time, to our knowledge, two such systems represent the technology most closely related to the present invention, those being patents 4551776 and 4578723.

『Automatic Reference Adjustment for Position Er
No. 4551776 entitled "Ror Signal on Disc File Servo System" periodically recalibrates the position error signal and increases the magnitude of the position error signal by modifying the gain of a variable gain amplifier in the servo loop. Do or reduce.

『Head Positioning System with Automatic Gain Co
Patent No. 4578723 entitled "ntrol" describes a system with automatic gain control. The system is substantially independent of head width, limiting the variation of off-track gain between heads. The position of the transducer head is controlled by the head position determining actuator means using the multiple phase radial position error signal obtained from the position reference information on the disk. The gain of the variable gain amplifier is responsive to a signal from a special magnetic head connected in the servo loop and is controlled by the gain function. The gain function is derived by combining different phase position error signals,
The rate of change of the position determination error signal is measured for each track displacement regardless of the position of the magnetic head. The output of the variable gain amplifier is used to control the position of the magnetic head.

Therefore, according to the teaching of the first patent, the position determination error signal is recalibrated by modifying the gain of the variable gain amplifier to keep the position within limits. According to the teaching of the second patent, the rate of change of the position determination error signal for each track displacement is used to control the gain of the variable gain amplifier.

[Object of the Invention]

The present invention has been made to solve the above-mentioned drawbacks.

[Outline of Invention]

The arrangement according to the present invention further improves the equalization of servo gain among a plurality of magnetic heads in a disk drive. In that arrangement, servo gain is measured when connected to each magnetic head at a different track position on each memory disk. Therefore, the amount of change in the servo gain between the heads and the amount of change in the servo gain for each head when traveling the related disk can be obtained. The correction of the servo gain, i.e. the reference at the selected track position, is determined. These are stored and called individually to select a magnetic head and its track position. The servo gain is essentially the same for each head as each head and track position is automatically compensated using the recalled servo gain reference, and each head moves across the memory disk surface. But it is kept within very small limits. Servo gain correction, a reference for each magnetic head, is stored on a memory disk and read by the head for that disk when the drive is powered up, or a separate memory, eg programmable for a particular disk drive. To read-only memory.

Example of Invention

Servo gain variation due to different magnetic heads connected in the servo loop presents a serious servo design problem. The electrical and temporal performance of the magnetic head changes. Therefore, not only are the servo gains of the heads connected to the servo loop different, but the servo gains of the individual heads also vary with the different radial positions of the magnetic heads on their respective memory disks. Figure 1 approximates this situation with a simplified graph, which shows the servo gain for two different heads at different track positions between the outer track (OT) and the inner track (IT) in decibels. It is plotted in. The servo gain changes are plotted linearly to simplify the figure. It is clear from this figure that the servo gain is different at any radial track position of the memory disk when the head is connected into the servo loop. It is desirable to limit the servo gain to a narrow band, for example between the limits of lines L1 and L2 of FIG. Then, the magnitude of the gain and the gain bandwidth are relatively constant over the entire disc track.

To this end, each head is moved to, for example, track positions T1, T2, T3 and T4 when connected in a servo loop and the servo gain is measured at each point. The servo gain at each point is used to generate a correction amount or a reference amount or signal for each head at each track position. They are used to keep the servo gain at each track position T1, T2, T3 and T4 within the acceptable limits shown by lines L1 and L2.

This result is shown in FIG. 2 for a single head, for example head H1.

Here, the corrected servo gain for head H1 at each point T1 to T4 is approximately halfway between lines L1 and L2. As the magnetic head moves across the track, the corrected rate of change in servo gain is close to the rate of change in the uncorrected plot of servo gain, but corrected at the midpoint between points T1 and T4 where the servo gain is measured. As a result, the sawtooth servo gain changes as shown in the figure. Such servo gain correction for each magnetic head results in an individual servo gain correction profile. The correction profile roughly corresponds to that shown in FIG. 2 and lies substantially within the boundaries of the lines L1 and L2. By this means, the servo gain for each head is substantially the same as each head is connected into the servo loop and remains substantially constant. That is, the servo gain is within limits defined over the range of track positions from the inner track to the outer track. Servo gain corrections are stored separately and are recalled each time the head and track are selected to compensate for servo gain.

Inventive environment A disk drive or disk file is usually
A plurality of memory disks are included, the memory disks being fixed along the axis of the shared spindle at regular intervals in the axial direction. The shared spindle is driven by a motor and drives the disk at a constant speed. Data is recorded on both sides of these discs, but only one side of one disc is assigned for servo codes. Each data track has at least one section of servo code, or sector. Hereinafter, it is called a sample servo. By the arrangement of the sample servo code,
The sector of each data disk surface is determined. There are several sectors around the face of each disc. There are as many sector marks on the dedicated servo surface as there are sectors on each data disk. The beginning of the sector mark on the dedicated servo surface and the beginning of the sample servo sector on the data disk surface are vertically aligned. One sector mark on the dedicated servo surface identifies the first sector from which the sectors are counted. The sector count on the dedicated servo disk serves as the sector count for all data disks.

Each magnetic head is held on a flexible arm at one end of the movable carriage. A servo drives the movable carriage. The servo moves the selected head to the selected track depending on the selection, i.e. the individual head requested and the track address. This is a seek mode operation. The seek operation proceeds under the control of the dedicated servo head on the dedicated servo surface.
When the seek movement is completed, track tracking is performed on the selected track by using the dedicated servo head first, and then using the selected head called the sample servo head. The sample servo head senses the sample servo code in a selected track on the relevant disk surface.

The sample servo gain corrections and dedicated servo gain corrections determined for each head at each track position T1-T4 are stored, eg, in a programmable read-only memory, in accordance with one embodiment of the present invention. The request from the host computer is processed as a servo gain correction memory address for the selected sample servo head on the selected track. The track address also invokes servo gain correction for the dedicated servo head. Addressed and stored corrections are multiplexed out of memory to create sample servo gain corrections and dedicated servo gain corrections, which are combined with the outputs from the sample head and dedicated head, respectively. Multiplexing is performed using the gain selection signal of FIG. 3 obtained from the dedicated servo sector mark. This is a square wave signal and switches between dedicated servo gain correction and sample servo gain correction depending on the respective voltage states. These voltage states are repeated for each sector. Sector marks provide a convenient source of gain selective multiplexing signals where sector marks are available on the disc. Other methods may be used to produce this signal that matches the circuit time constant.

A sixth servo system embodying and implementing the present invention
Shown in the block diagram of the figure. Two separate control systems with a common part are used to control the position of the head. One is a linear position control system, that is, a fine adjustment position servo, which is used for a track following operation mode and holds a dedicated servo head or a sample servo head selected at the center of the servo track. This is an analog position control system. The other is a non-linear position control system used to move the head from one track to another during seek mode operation.

Only three heads are shown in FIG. One is a dedicated servo head DH corresponding to a dedicated servo code, and the remaining two, H1 and H2, are sample servo heads, each of which reads or writes data on the data surface of the corresponding data disk. It is also used for the purpose of track following using the sample servo code.

The linear position control system of FIG. 6 is used at the end of the seek motion to center the dedicated head first, and then the selected sample servo head for reading or writing data. The linear system consists of a sample servo track follower 10, or a dedicated servo track follower 12, which is selectively connected to the position compensation stage 14 by a switch S1. The position compensation stage consists of a compensator 16 and a feedforward network 18. The output of compensation stage 14 passes through switch S2 and summing junction 22 when switch S2 is closed and is connected to filtering and amplification network 20. The filtering and amplification network 20 typically comprises low pass filters, notch filters and power amplifiers (all not shown). Filter amplifier 20 is an actuator
The actuator 24 controls the carriage 24, and the actuator 24 drives the carriage 26. The carriage 26 is connected to the head arm stack 28, and the head arm stack 28 moves the heads DH, H1 and H2 at the same time. The actuator 24 may be a magnetic driver having a movable member that drives the arm stack carriage 26. A head arm stack assembly 28 is attached to the arm stack carriage 26. This closes the position loop. The input circuit of the sample servo track follower 10 is selectively connected to the head H1 or the head H2 by a switch circuit (illustrated as a switch S3 here).

In seek operation mode, switch S2 opens. Requests from host computer 30 processed by drive controller 32 are connected to servo processor 34 by controller interface 36 as the distance to go. When the switch S2 is opened, the servo processor controls the seek operation. Produced by the servo processor,
Use a negative or positive acceleration command current through the circuit NC or PC to drive the actuator
ng) Driven by type control. Non-linear position control systems are used to move long movements, i.e., across one or more tracks. The travel distance determined by the target address is in this case the track count but always known before the seek movement begins. The servo processor 32 is a Model 8051 manufactured by Intel Corporation of Santa Clara, California, although others can be used. There are no velocity transducers in this system. A servo processor is used to control the movement of actuator 24 based solely on the track crossing information provided by track crossing detector 38. With bang-bang servo control, the acceleration commanded to actuator 24 is either on or off.

There are two types of servo movement in the seek operation mode. One is an open loop move and the other is a closed loop move. During a seek operation, the open loop system accelerates and decelerates the head to the final velocity at which the closed loop system is acceptable, and the final position. Then, in the track following operation mode, the closed loop system functions to follow the selected head. At the end of the seek operation, switch S2 is closed and switch S1 connects the dedicated track follower 12 to the position compensation stage 14. In this track following operation mode for the dedicated head DH, the dedicated head follows the center of the selected track. When the dedicated head moves and the centering of the track is completed, the switch S1 connects the sample track follower 10 to the position compensation stage 14. At this time, the switch S3 moves to connect the selected sample servo head H2 to the sample track follower 10.
Here, the sample servo head H2 is aligned with the center on the selected track.

Seek operation mode and track following operation mode
It is under the control of the drive controller 32 in response to the request of the host computer 30. Therefore, in response to a request from the host computer 30, for example, the selected head H1
Alternatively, a head address signal is sent to the drive controller 40 via the circuit 33 for the read or write operation of H2. Drive controller 40 is switch S3
Control. If the head address selects the head H2, the switch S3 connects the head H2 to the sample servo track follower 10 again. This information is also passed through the controller interface 36 to the servo processor.
Conveyed to 34. Here, the servo processor controls the dedicated track follower 12 to start the seek operation mode,
Switch S2 is opened and the dedicated servo track follower 12 is connected to the position compensation network 14 by means of switch S1 at some point in the cycle of the seek operation. In the seek mode of operation, the track crossing detector 38 provides the servo processor with the number of actual track crossings during a move to a new track address, and the analog-to-digital converter on the selected track provides the track address. The controller interface 36 compares the track address with the track address supplied to the servo processor.

Seek mode is completed with the correct track address.
The servo processor closes switch S2 and closes the fine position servo loop. Here, the output of the dedicated track follower connected to the position compensation network 14 through the switch S1 is input to the servo, and the dedicated head is aligned with the center of the selected track. Once the dedicated head is centered on the track, the servo processor actuates switch S1 to connect the sample servo track follower 10 to the position compensation network 14. The selected sample servo head H2 is now in the servo loop and is centered on the selected track. The off-track detector 44 is used for disabling a read operation or a write operation. If the selected head gets lost off the track, the output of the off-track detector sends a signal to the servo processor to disable operation.

Servo Gain Compensation The gain of a servo system varies from head to head and also inside and outside the disk track. This gain change must be minimized to maintain the maximum margin in the servo system. This is done simultaneously for both the sample servo and dedicated servo functions. As mentioned above, the servo gain for each magnetic head is measured at four track positions across the disk surface.
The signal gain correction required to equalize the servo gains is then calculated and stored in programmable read-only memory. The drive controller 32 calls it. The servo gain compensation system is shown in FIG. Here, the servo gain compensation network 46 is the track follower circuit 10
And 12 and the drive controller 32. As will be described below, the servo gain compensation network 46 includes a programmable read only memory in which servo gain corrections for each head at each track position are stored. The head and track addresses are connected to programmable read-only memory in servo gain compensation network 46 by circuits 33 and 35 from the drive controller. Circuits 37 and 39 respectively connect the sample servo gain correction signal and the dedicated servo gain correction signal to the sample track follower 10 and the dedicated track follower 12, respectively. Since the servo gain compensation network 46 has only one programmable read-only memory, the gain selection signal of FIG. Switch the servo gain correction current with. By this means, the output of each track follower keeps the servo gain within acceptable limits. That is, the margin is maintained at the maximum in the servo system.

Details of the servo gain compensation network are shown in FIGS.
Shown in the figure, in which FIG. 6 is a block diagram of a compensation network. Referring now to FIGS. 6-9, drive controller 32 provides head address inputs and track address inputs to programmable read-only memory 48. The servo gain correction signal is stored at the track address for each head as described above. These track and head addresses are in memory
Although shown in more detail in FIG. 8 as an input to 48, some memory configurations may include a memory chip address as shown. The gain select signal is also an input signal to memory 48. The 8-bit digital output from programmable read-only memory 48 is input to digital-to-analog converter 50.

The digital / analog converter 50 is conventional and produces an analog servo gain correction current SGC in response to an 8-bit input. The analog servo gain correction current SGC is a constant multiple of the reference current DAR shown in FIG. The servo gain correction current from the digital / analog converter creates a voltage across a resistor 51 connected between the output of the digital / analog converter 50 and ground. When only the most significant bit of the normally programmable read-only memory 48 is set, there is a constant voltage across this resistor,
For example, -1 volt is output. As the current from the digital-to-analog converter changes, the voltage across resistor 51 changes by a fraction of a volt. For example, a 1/128 volt change per bit of the binary value of the servo gain correction determined by the contents of programmable read only memory 48.

Since both servo gain corrections need to be maintained at the same time, the programmable read-only memory and the output of the digital-to-analog converter are stored by switching by the gain select signal means with a selected sample gain correction value and a dedicated automatic gain correction. Multiplexed with the value. The field effect transistor switch 52 connected to the output of the digital to analog converter is switched at the same time as the address of the read only memory 48 programmable by the gain select signal. The servo gain correction current SGC is connected to the switch 52. FIG. 7 is mechanically equivalent to a field effect transistor switch. Storage capacitor 54
And 56 are connected to the respective output circuits of switch 52 and thus temporarily store the sample servo gain correction and the dedicated servo gain correction. Therefore, the switch frequency is selected to match the time constant of the capacitor circuit under the control of the gain select signal. Capacitors 54 and 56 are buffered by buffer amplifiers 58 and 60 respectively so that the capacitor voltage does not drop until the next sample. The gain select signal may be conveniently produced by a dedicated servo head in response to the start of each sector of the dedicated servo surface.

If a failure occurs, the gain select signal will automatically turn off as in the off track condition and the automatic gain control multiplexing system will remain in the dedicated servo mode. This is necessary to get the system back on track. A resistor 53 connected between resistor 51 and capacitor 54 ensures that the charge on capacitor 54 does not drift away from its nominal gain value, even if this condition exists for some time.

The output of the sample servo buffer amplifier 58 is the reference voltage for the sample servo AGC integrator 62 and is also multiplied by a negative constant by the ground referenced amplifier 64. Amplifier 64
The output of is the sampled servo correction reference SSCR. This is the reference voltage for the sample servo dibit detector 66. Sample servo correction reference voltage SSCR
Is maintained at a constant rate of signal amplitude. Similarly, the output of the dedicated servo buffer amplifier 60 is the reference voltage for the dedicated servo automatic gain control integrator 68, and the output of the dedicated servo automatic gain control integrator 68 is the dedicated automatic gain control reference DAGC. The output of the dedicated servo buffer amplifier 60 is also multiplied in amplifier 70 by a negative factor which is referenced to ground,
It produces an output ZDV which is a dedicated servo reference voltage. The dedicated servo reference voltage is connected to the dedicated servo dibit detector 72. The other input 67 and 73 to each dibit detector 66 and 72, respectively, is the filtered and amplified sample servo and dedicated servo head voltage. In the case of the sample servo dibit detector, this is the filtered and amplified voltage from the selected sample servo head. The output of the dibit detector is the servo signal corrected for servo gain. Although the dibit detectors are shown separately from their respective track follower circuits, they are considered part of the sample servo track follower circuit 10 and dedicated servo track follower circuit 12 of FIG. The output of each track follower circuit 10 and 12 is connected to the switch S1 as shown in FIG.

In the track following mode, the track centering signal is generated from the magnetic transition of the servo code. The magnetic transitions of the servo codes are 180 degrees out of phase with each other within one period on the disk surface, and determine, for example, the opposite side of the data track or the dedicated servo track. If these generated signals are equal, the head is considered to be centered on the track. If one or the other of these signals is large, the head moves radially on the track and the servo responds so that the signal amplitudes are equal. In such a device, amplitude control or amplitude limitation is very important for realizing proper servo operation. Therefore, the reference voltages V _SLVL and V _DLVL connected to the positive terminals of the automatic gain control integrators 62 and 68 are the sum of the two voltages obtained from the magnetic transitions of the servo codes that determine the end of each track. Derived from. Therefore,
If these voltages are designated as voltages A and B in FIG. 9, the sum of these two voltages is the sample servo level voltage in the case of the sample servo and the dedicated servo level voltage in the case of the dedicated servo. These are connected to the negative terminals of each of integrators 62 and 68. The characteristics of the voltages A and B for the head aligned with the track center are shown in FIG.

FIG. 9 shows a circuit for generating the sample servo level voltage A + B. These voltages, derived from the magnetic transitions of the servo code on a particular data track by the sample servo head, are connected as inputs to a variable gain amplifier 74. The gain of the variable gain amplifier 74 is controlled by the sample servo automatic gain control voltage SAGC from the amplifier 62. The output A + B represents the sample servo level voltage V _SLV . This output passes through a sample position demodulator 75 and an amplifier 62.
Is connected to the negative input terminal. This circuit is also part of the sample servo track follower 10 of FIG. The other output AB of demodulator 75 is the track following signal connected to switch S1. Dedicated track follower circuit 1 for producing a dedicated servo level voltage V _DLVL for amplifier 68
The same circuit applies to 2.

In the device discussed, once the servo gain correction has been determined and stored in programmable read-only memory 48, this memory is used by the head disk assembly (HD
Must be retained with A). The correction is unique to the HDA. One way to ensure that the memory is with the HDA is to physically lock the memory to the HDA. This is accomplished by physically fixing the memory as well as a digital / analog converter connected to a flexible circuit that connects the magnetic head to the drive electronics. In this case, memory and digital /
An analog converter is conveniently connected to the servo system through a flexible circuit connector.

Alternatively, the servo gain correction for the disk drive may be stored on one of the memory disks. With this method, the servo gain correction for the disk drive becomes inseparable from the HDA. The correction may be written anywhere convenient on the disc. For example, the track position on the inner circumference,
Alternatively, it may be either the track position on the outer circumference. Such an arrangement replaces programmable read-only memory 48 in the system with random access memory. This random access memory is with the electronics and is not a permanent part of the HDA like programmable read only memory 48.

Servo gain correction can be written on any of the memory disks. In a system such as that shown in FIG. 4, the correction may be written on the dedicated servo code plane. However, whether or not dedicated servo control is used for head position determination, the correction can be written to any other disk.

2 for servo gain correction using a magnetic disk for storage
Two approaches are used. One uses a preset servo gain to read all servo gain corrections recorded on a single track on the disk surface, and the other uses the inner or outer crash stop.
Align the position of the head carriage to p) and record servo gain correction on many tracks here as well.

In the first approach, the servo gain is calibrated or approximated for a selected head at a constant radial position. Resistors in the servo system can be used for this purpose. The preset gain is then used to write servo gain corrections for all other heads to signal tracks at all convenient locations on the disk. The correction is thus permanently attached to the HDA, along with a means for invoking the correction.

The selected head and track are recalled whenever the disk drive is powered on. The servo gain correction of the selected track can be read by the preset servo gain. In track seeking and track following operations, the read corrections, whether using dedicated servos or sample servo techniques, are stored in random access memory for use in the servo system.

In the second approach, servo gain corrections are recorded on the inner or outer radius of the memory disk. Here, the same recording is performed on many adjacent tracks for rough alignment of the head used when reading the correction. Whenever the disk drive powers up, take the carriage to the crash limit of the selected radial limit. The servo gain correction is read and stored from random access memory for use in the servo system as described above.

Although specific circuits and specific examples are shown herein to illustrate the best mode of carrying out the invention, other circuits and techniques may be used to implement the invention.

〔The invention's effect〕

As described above, by using the present invention, it is possible to minimize the servo gain change due to each head or the difference between the inner track and the outer track, and thus it is possible to maintain the maximum margin in the servo system.

[Brief description of drawings]

FIG. 1 is a diagram depicting a change in servo gain between different magnetic heads and a change in servo gain of heads connected to different track positions on a memory disk, and FIG. 2 is a magnetic head at different track positions on a memory disk. FIG. 3 depicts one technique of the present invention for keeping the servo gain of an essentially constant servo; FIG. 3 shows the signals used to switch between a dedicated servo gain calibration and a sampled servo gain calibration; FIG. 4 is a block diagram of a preferred embodiment of the present invention, FIG. 5 is a diagram depicting a typical servo speed profile, and FIG. 6 is a block diagram showing features of servo gain compensation of one embodiment of the present invention. FIG. 6a is a diagram showing a second embodiment of the present invention which is a modification of FIG. 6, FIG. 7 is a diagram depicting the function of the field effect transistor switch of FIG. 6, and FIG. 8 is a programmable diagram of FIG. Read only Further detailed development block diagram of the memory and the digital / analog converter, FIG. 9 is a diagram showing a circuit for generating the reference and level voltages for the sampled servo AGC integrator and for generating the track position error voltage. . 10: Sample servo track follower 12: Dedicated servo track follower 32: Drive controller 34: Servo processor 38: Track crossing detector 46: Servo gain compensation network

Continuation of the front page (56) Reference JP-A-56-134367 (JP, A) JP-A-60-66373 (JP, A) JP-A-58-215767 (JP, A)

Claims (1)

[Claims] 1. A dedicated servo disk having a dedicated servo code, a dedicated servo magnetic head associated with the disk, a plurality of memory disks having a sample servo code for each track, and a sample servo magnetic head associated with the disk. A magnetic disk drive comprising means for selectively connecting a magnetic head associated with each disk to a servo loop and means for positioning the magnetic head at a predetermined track on the associated disk via the loop , Each of the disks has a plurality of predetermined tracks corresponding to each other, which are intermittently selected on each disk, and obtains a servo gain for each of the predetermined tracks with respect to the selected magnetic head. A means for determining a servo gain compensation value for each of the predetermined tracks based on a gain; Corresponding to a storage means for storing a compensation value, a servo gain compensation value corresponding to a combination of the magnetic head and the track when the sample servo magnetic head and the track are selected, and a corresponding track on the servo disk. And a servo gain compensation value which is read from the storage means in a time division manner to compensate each servo gain.